2.0 Components of a Wet Weather Crash Reduction Program - State Practices to Reduce Wet Weather Skidding Crashes

<< Section 1.0 – Introduction and Background

Table of Contents

Section 3.0 – State Practices to Reduce Wet Weather Skidding Crashes >>

2.0 Components of a Wet Weather Crash Reduction Program

The review of state wet weather crash reduction programs identified four common program components: identification of wet pavement crash locations, friction testing procedures, investigation and remedial action of wet pavement crash locations, and project and program evaluations. This section describes the importance of each component and related FHWA guidance.

2.1 Identification of Wet Weather Crash Locations

The identification of locations with a high frequency or proportion of wet pavement crashes is a key component of a wet weather crash reduction program. The FHWA Technical Advisory on Pavement Friction Management [2] provides three common approaches for agencies to use to analyze the data in the state crash database to identify wet weather crash locations:

• Identify locations with a wet crash ratio above a specified value as a high wet weather crash location. (The ratio of the wet weather crashes to total (wet+dry) crashes.) The specified value varies between agencies depending on geometric and climatic circumstances; typically the ratio varies between 0.25 and 0.50.
• Identify locations with a wet crash ratio above the average wet crash ratio for the corresponding functional classification of highways. If a location is above the average by a specified percentage, the location is identified as a wet weather crash location.
• Identify locations that exceed an established wet crash ratio and a minimum number of wet weather crashes within a specified segment length as a wet weather crash location. As an example, one agency uses a minimum of six wet road crashes in rural areas and a minimum of 10 in urban areas.

Segment lengths used to compute wet crash ratios vary by agency, but typically a segment length of 0.2 to 2.0 miles is used. Once sites are initially identified through an analysis of the crash database, the sites should be friction tested and further investigated for potential remedial action.

2.2 Friction Testing Procedures

Pavement friction testing is an integral component of any wet weather crash reduction program focused on skidding crashes. Typically, as pavement friction decreases, the number of wet weather crashes will increase. The FHWA Technical Advisory on Pavement Friction Management [2] provides guidance to state and local agencies in managing pavement surface friction.

Two types of surface texture affect wet pavement friction: microtexture (wavelengths of 1µm to 0.5mm) and macrotexture (wavelengths of 0.5mm to 50mm). Microtexture is generally provided in asphaltic pavements by the relative roughness of the aggregate particles and in concrete surfaces by the fine aggregate. Macrotexture is generally provided in asphalt pavement by proper aggregate gradation and in concrete surfaces by a supplemental treatment such as diamond grinding or grooving, exposed aggregate texture, transverse or longitudinal tining, burlap or artificial turf dragging, and transverse brooming.

Friction Testing Methods

Four types of full-scale test equipment exist for measuring pavement friction, including locked wheel, fixed slip, side force, and variable slip. However, the recommended methods for evaluating pavement friction on U.S. highways are the locked wheel and fixed slip methods; currently side force and variable slip friction measurement systems are not widely available or used in the U.S. Table 2.1 identifies the advantages of each of these four testing methods.

Table 2.1 Advantages of Friction Testing Methods

Method	Advantage
Locked wheel (ASTM E 274)	Simulates emergency braking without anti-lock brakes. Can be used with either the ribbed tire (ASTM E 501) or the smooth tire (ASTM E 524).
Fixed slip	Relates to braking with anti-lock brakes. Ability to operate continuously over a test section.
Side force	Measures the ability to maintain control on curves.
Variable slip	Relates to braking with anti-lock brakes

Source: FHWA Technical Advisory on Pavement Friction Management [2].

The ribbed tire (ASTM E 501) is the most common test tire used by U.S. state highway agencies with the locked wheel method, but it is considered less sensitive to pavement macrotexture and water film depth compared to the smooth tire (ASTM E 524). However, all friction test methods can be insensitive to macrotexture under specific circumstances, so it is recommended that friction testing be complemented by a macrotexture measurement (ASTM E 1845). Macrotexture measurements can be independently used to compute the Speed Gradient (Sp). (Sp defines the relationship between measured friction and vehicle tire free rotation or slip speed.) The Sp can then be combined with friction results from most friction testers to determine the International Friction Index (IFI). The IFI can be used to directly compare friction test results using different test methods. The AASHTO Guide to Pavement Friction [3] provides models to use for these conversions.

To ensure reliable friction test results, it is essential to proper calibrate and maintain friction testing equipment as specified by the manufacturer.

Friction Testing Conditions

Friction test results can be impacted by various factors such as surface temperature, test speed, and ambient weather conditions. Conducting friction testing under standardized conditions helps to minimize the effects of these factors, which minimizes variability and produces repeatable measurements. Table 2.2 summarizes AASHTO’s guidance on standardized test conditions [3].

Table 2.2 Standardized Test Conditions

Factors	Consideration
Season	Limit friction testing to a specific season or time of year when friction is typically the lowest to maintain year-to-year consistency and reduce variability in measured data. When this is not possible, correction factors can be developed to normalize raw friction test data to a common baseline season or coordinate initial and subsequent section testing to occur during a specific season.
Test speed	The standard speed recommended by AASHTO T 242 for pavement friction tests (locked wheel) is 40 miles per hour. However, since most agencies conduct friction tests without traffic control and posted or operational speeds vary throughout the roadway network, it is difficult for the operator to conduct testing at just this speed. Therefore, results of friction testing conducted at speeds other than 40 miles per hour should be adjusted to the 40 miles per hour baseline to make friction measurements comparable. This requires establishing correlations between friction measurements taken at 40 miles per hour and those taken at other speeds.
Test lane and line	Friction testing must be done in the most heavily trafficked lane, since it expected to have the highest rate of friction loss due to wear. Two-lane highways with a near 50-50 directional distribution of traffic only need to be tested in one lane; otherwise, the lane direction with the higher traffic volumes should be tested. For multilane highways, the outermost lane in both directions is typically the most heavily traveled and should be tested. Testing must be carried out within the wheel path, since this is the location where friction loss is the greatest. Testing should be carried out in the same lane and wheel path to maintain consistency between test results and reduce variability. If it is necessary to deviate from typically practice, the test data should be marked accordingly.
Ambient conditions	It is important to standardize and document ambient test conditions, as they can have an effect on friction test results. Avoid testing in extremely strong side winds as these can create turbulence under the vehicle causing the water jet to divert from the correct line and affect the measurement results. Avoid testing in heavy rainfall or where there is standing water on the pavement surface. Excess water on the surface can affect the drag forces at the pavement-tire interface and influence the measurements. Do not conduct testing when the air temperature is below 41ºF (5ºC).
Contamination	Avoid testing locations where the pavement surface is contaminated by mud, oil, grit, or other contaminants.

Source: AASHTO Guide to Pavement Friction [3].

Establishing Friction Thresholds

There is not a specific friction test value that represents the difference between a safe and potentially unsafe pavement surface. Each agency determines their own investigatory (or desired) friction levels or friction-level ranges for specific facility types, based upon factors such as traffic volume, geometrics (e.g., curves, grades, sight distance), potential conflicting vehicle movements, speed, and intersections. Once sites fall into the investigatory friction-level range, they are further investigated. Many states also develop intervention friction-level thresholds that represent a minimum level of pavement friction. Once sites reach the intervention level, some type of action is required. These thresholds can help an agency in prioritizing improvement projects for sites identified as wet crash locations.

The AASHTO Guide for Pavement Friction [3] identifies three methods for establishing investigatory and intervention-level friction thresholds. The first method establishes thresholds by examining historical pavement friction data to determine at what pavement age significant decreases in friction occur and set thresholds based on those friction values. The second method compares historical friction and crash data and establishes an investigatory level based on large changes in friction loss and an intervention level based on when there is a significant increase in crashes. Finally, the third method establishes thresholds based on friction distribution and crash rate.

Resurfacing Projects

A program targeting the reduction of wet weather crashes through surface friction improvements, should give consideration to the resulting friction of the surface treatment. Friction testing and friction-related specifications on new hot-mix asphalt or concrete surfaces may be justified unless historical evidence indicates that existing pavement mix-design requirements, aggregate specifications, or construction specifications have resulted in pavement surfaces that provide adequate pavement friction.

2.3 Investigation and Remedial Action of Wet Weather Crash Locations

Sites identified during the crash data analysis need to be further investigated to determine potential contributing factors to the crashes. The friction number is evaluated as part of this investigation. Typically if the friction number falls below the investigatory threshold, the site is reviewed in the field to identify existing conditions and determine potential contributing factors and potential improvements. This investigation also may include a detailed analysis of the individual crash reports to identify collision patterns. The field review will also identify any site conditions that may have contributed to the crashes, including potential cross-section and pavement deficiencies. If low pavement friction is identified as a contributing factor, the next step is to identify the appropriate remedial action. The FHWA Technical Advisory on Surface Texture for Asphalt and Concrete Pavements [4] identifies several techniques to provide adequate surface friction on new pavements and overlays and to restore surface friction of existing pavements. These techniques are summarized in Table 2.3.

Table 2.3 Techniques to Provide Adequate Surface Friction

Technique	Description
Concrete Surfaces
Transverse tining	Achieved by a mechanical device equipped with a tining head that moves laterally across the width of the pavement surface. A width of 3mm (±0.5mm) and a maximum depth of 3mm are recommended. Recommended random spacing average tine spacing of either 13 mm with a tine spacing pattern of 10/14/16/11/10/13/15/16/11/10/21/13/10 mm or 26 mm with a time spacing pattern of 24/27/23/31/21/34 mm.
Longitudinal tining	Achieved by a mechanical device equipped with a tining head that moves parallel to the pavement centerline. A width of 3mm (±0.5mm) and a maximum depth of 3mm are recommended. Narrower, deeper grooves are better than wider, shallower grooves (within the limits) for minimizing noise. Straight, uniformly spaced grooves at 19mm have been shown to provide adequate handing characteristics for small vehicles and motorcycles.
Exposed aggregate	Normally constructed in two layers. The top layer consists of 30% siliceous sand of 0-1 mm and 70% high-quality chips of 4-8 mm. A water cement ratio of 0.38 and a mean texture depth of 0.77 mm are recommended.
Diamond grinding	Typically provides grooves of approximately 3mm width, spaced at 5‑6 mm intervals. Specific groove depth and spacing is dependent on hardness of aggregate.
Diamond grooving	Transverse or longitudinal can provide adequate friction characteristics. Groove geometry should be consistent with recommendations for tining.
Burlap drag	Typically produced by trailing a moistened, course burlap from a construction bridge that spans the pavement. Striations of 1.5‑3 mm depth are typical.
Artificial turf drag	Typically produced by trailing an inverted section of artificial turf from a construction bridge that spans the pavement. Striations of 1.5‑3 mm depth are typical when using turf with 77,500 blades per square meter.
Transverse broom	Typically obtained using a hand broom or mechanical broom device that lightly drags stiff bristles across the surface. Striations of 1.5‑3 mm depth are typical.
Thin epoxy laminates	Typically aggregates are 4‑6 mm.
Asphalt-based surface treatments	May include micro-surfacing.
Asphalt surfaces
Surface treatments or thin asphalt overlays	Generally, hot-mixed asphalt pavements designed in conformance with Superpave mix design will provide adequate macrotexture and microtexture without supplemental treatments. When supplies of durable nonpolishing aggregate are limited, an agency may choose to construct an asphalt pavement using high-durability aggregates optimized for friction properties only in the top layer.
Concrete overlays	May be considered as an option to restore adequate surface texture on asphalt pavements.

Source: FHWA Technical Advisory Surface Texture for Asphalt and Concrete Pavements [4].

There are several different techniques for improving pavement friction, and due to widely varying conditions of different sites, it is unlikely that one texturing method will be the optimal choice for all projects within a state. The selection of the appropriate technique should consider the existing conditions at each individual site. FHWA [4] identified several factors to consider when selecting a method to improve pavement friction, including:

• Splash and Spray – Reduced visibility caused by splash and spray may increase the probability of wet-weather crashes. Adequate pavement cross-slope or the use of porous surfaces will provide improved surface drainage and has been shown to reduce splash and spray.
• Climate – The increased probability of wet-weather conditions would justify a higher level of texture.
• Traffic Volume and Composition – Pavements with higher traffic volumes can justify a higher level of texture. Increased traffic would decrease the reaction/recovery time in the event of loss of control of a vehicle. Additionally, roadways with a higher composition of truck traffic typically demand a higher level of friction compared to corresponding highways comprised predominately of passenger cars.
• Speed Limit – Higher speed facilities may justify a higher level of texture. Friction test results will decrease with increasing speed, reaching a minimum at approximately 60 mph. Friction on surfaces with low texture falls more rapidly with speed than on high-textured surfaces.
• Roadway Geometry – Research has shown that curves tend to lose pavement friction at a faster rate than other roadway locations, and therefore, curves may justify a higher level of texture.
• Potential Conflicting Movements or Maneuvers (Frictional Demand) – Intersections and presence of pedestrians will justify a higher level of texture due to the increased likelihood of sudden braking movements.
• Materials Quality and Cost – The availability and cost of high-quality durable, nonpolishing materials will influence the choice of materials and techniques to provide increased friction.
• Presence of Noise-Sensitive Receptors – A pavement located near a school, hospital, or other noise-sensitive receptor may justify a higher consideration of noise effects when selecting the appropriate surface treatment for a pavement.

New and/or innovative pavement friction improvement techniques, without evidence of improved safety performance, should only be used on an experimental basis and monitored for safety performance.

2.4 Project and Program Evaluations

The purpose of a wet weather skidding crash reduction program is to reduce the number of crashes occurring on wet pavement due to inadequate pavement friction. Conducting project and program evaluations enables an agency to determine if their efforts have met their intended purpose and provides a quantified measure of success. Individual projects can be evaluated based on the occurrence of wet pavement crashes before and after an improvement. States evaluate safety projects implemented for the Highway Safety Improvement Program (HSIP) on an annual basis, and these same methods can be used to evaluate the effectiveness of pavement improvements at wet weather crash locations.

As part of the Technical Advisory on Pavement Friction Management [2], FHWA identified the “wet safety factor” (WSF) as an appropriate metric for evaluating the effectiveness of a wet weather crash reduction factor. The WSF is the reciprocal of the risk of having a wet pavement crash relative to a dry pavement crash and is calculated as follows:

Where,

DC = Number of dry weather crashes;

WC = Number of wet weather crashes;

PDT = Percent of dry pavement time; and

PWT = Percent of wet pavement time.

To determine a composite statewide WSF, the network is divided into analysis areas based on similar percentages of wet and dry pavement time. The total number of dry and wet weather crashes are determined for each analysis area and used to calculate a WSF for each analysis area. Then, the WSF for each analysis area is weighted by the vehicle miles traveled (VMT) and aggregated to determine a composite statewide WSF. In a successful program one would expect the WSF to increase over time with an upper limit of 1.0. A WSF less than 0.67 suggests a potential wet weather problem. This value is based on the conservative estimate of the overall likelihood of a wet weather crash being 1.5 times greater than a dry pavement crash.

<< Section 1.0 – Introduction and Background

Table of Contents

Section 3.0 – State Practices to Reduce Wet Weather Skidding Crashes >>